CtrlK
BlogDocsLog inGet started
Tessl Logo

claude-code-debug

Troubleshoot Claude Code extensions and behavior. Triggers on: debug, troubleshoot, not working, skill not loading, hook not running, agent not found.

Invalid
This skill can't be scored yet
Validation errors are blocking scoring. Review and fix them to unlock Quality, Impact and Security scores. See what needs fixing →
SKILL.md
Quality
Evals
Security

Evaluation results

70%

22%

Audit Failing Claude Code Hook

Hook debugging and validation

Criteria
Without context
With context

chmod +x recommendation

100%

100%

Manual hook test with JSON

100%

100%

Exit code check

0%

100%

jq JSON validation

0%

100%

/hooks command

0%

0%

Debug mode mention

0%

0%

Identifies root cause

100%

100%

Settings.json location

0%

0%

Structured diagnostic report

100%

100%

Without context: $0.3127 · 1m 38s · 14 turns · 21 in / 5,734 out tokens

With context: $0.7022 · 2m 50s · 32 turns · 148 in / 10,291 out tokens

94%

8%

Diagnose Non-Loading Claude Code Skill

Skill structure diagnosis

Criteria
Without context
With context

SKILL.md existence check

90%

100%

YAML frontmatter delimiters

100%

100%

Name matches directory

100%

100%

Trigger keywords check

25%

50%

Identifies name mismatch

100%

100%

Identifies missing delimiter

100%

100%

Fix applied to SKILL.md

100%

100%

ls directory check

60%

100%

Written diagnostic report

100%

100%

Without context: $0.5105 · 2m 33s · 23 turns · 29 in / 9,079 out tokens

With context: $0.6185 · 2m 19s · 30 turns · 38 in / 8,260 out tokens

86%

12%

Fix Ignored Custom Code Review Agent

Agent auto-delegation setup

Criteria
Without context
With context

Check both agent locations

25%

50%

"Use proactively" or "Use for:" addition

100%

100%

Description keyword inspection

100%

100%

claude --debug recommendation

0%

0%

Identifies missing delegation phrase

100%

100%

Explicit invocation fallback

100%

100%

Fixed agent file saved

100%

100%

/agents command

0%

100%

Written diagnostic report

100%

100%

Without context: $0.1836 · 1m 3s · 11 turns · 16 in / 3,389 out tokens

With context: $0.3538 · 1m 26s · 19 turns · 21 in / 4,610 out tokens

Repository
NeverSight/skills_feed
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.